Background/objectives: Opioids relieve symptoms in terminal care. We studied opioid underuse in long-term care facilities, defined as residents without opioid prescription despite pain and/or dyspnoea, 3 days prior to death. Design and setting: In a proportionally stratified randomly selected sample of long-term care facilities in six European Union countries, nurses and long-term care facility management completed structured after-death questionnaires within 3 months of residents’ death. Measurements: Nurses assessed pain/dyspnoea with Comfort Assessment in Dying with Dementia scale and checked opioid prescription by chart review. We estimated opioid underuse per country and per symptom and calculated associations of opioid underuse by multilevel, multivariable analysis. Results: Nurses’ response rate was 81.6%, 95.7% for managers. Of 901 deceased residents with pain/dyspnoea reported in the last week, 10.6% had dyspnoea, 34.4% had pain and 55.0% had both symptoms. Opioid underuse per country was 19.2% (95% confidence interval: 12.9–27.2) in the Netherlands, 25.2% (18.3–33.6) in Belgium, 29.3% (16.9–45.8) in England, 33.7% (26.2–42.2) in Finland, 64.6% (52.0–75.4) in Italy and 79.1% (71.2–85.3) in Poland (p < 0.001). Opioid underuse was 57.2% (33.0–78.4) for dyspnoea, 41.2% (95% confidence interval: 21.9–63.8) for pain and 37.4% (19.4–59.6) for both symptoms (p = 0.013). Odds of opioid underuse were lower (odds ratio: 0.33; 95% confidence interval: 0.20–0.54) when pain was assessed. Conclusion: Opioid underuse differs between countries. Pain and dyspnoea should be formally assessed at the end-of-life and taken into account in physicians orders.